Appl Microbiol Biotechnol (1991)36:1-4 017575989100231F Applied

Mrtcrobiology Bioteehnology © Springer-Verlag 1991

Mini-review

On the safety of Bacillus subtilis and B. amyloliquefaciens: a review

Anne Sietske de Boer and B~rge Diderichsen

Novo Nordisk A/S, Novo AII~, DK-2880 Bagsvaerd, Denmark

Received 28 May 1991/Accepted 6 June 1991

Introduction

For many years the fermentation industry has used mi- croorganisms to produce antibiotics, amino acids, en- zymes and other useful compounds. These microorgan- isms, which have been isolated from the environment and then mutated to increase yields of the desired prod- uct, have proved safe to handle. With the advent of gene technology, it is now possible to transfer genetic properties from one organism to another. It is widely accepted that as long as the recipient microorganism (the host) is harmless and the products of the genes to be transferred are innocuous, the genetically engi- neered microorganism (the recombinant) is as safe as the host.

An overwhelming majority of recombinant microor- ganisms to be used by industry are expected to be based on harmless hosts (OECD 1986). Many of these have been proven safe over many years of experience in industrial settings. Furthermore, extensive information on the incapacity to cause disease, i.e. non-pathogenic and non-toxicogenic potential, of some of these organ- isms can be found in the literature.

We believe that a review of the literature and present experience with some of these host organisms will be useful for assessment of the safety of many recombi- nant organisms. In particular it may help to classify some of these as GILSP (Good Industrial Large Scale Practice) host organisms as defined by the OECD (1986), thus facilitating the use of recombinant strains by established production procedures. Furthermore, sa- fety reviews on selected host microorganisms may ease the approval process of products produced by recombi- nant strains derived from these hosts. Thus it is the opinion of qualified experts that the use of genetic en- gineering per se does not warrant any additional safety assessment. On the contrary, use of a safe and well- known host organism may sometimes render super- fluous some of the extensive animal testing of a new

Offprint requests to: B. Diderichsen

product. For a more extensive discussion of the safety and regulatory aspects of the use of recombinant organ- isms see, for example, AMFEP (1990), Diderichsen et al. (1990), IFBC (1990), National Academy of Sciences (1987), and Pariza and Foster (1983).

Taxonomy and ecology

Bacillus subtilis is a Gram-positive, spore-forming bac- terium. It is commonly found in soil and on plant mate- rial and grows aerobically at intermediate temperatures and pH. As with many other bacilli, B. subtilis secretes substantial amounts of protein, especially hydrolytic enzymes such as amylases and proteases. B. subtilis is often referred to as a non-pathogenic bacterium and it is even consumed by humans in large quantities in the Japanese food natto (Djien and Hesseltine 1979).

The genetics of B. subtilis strain 168 (Burkholder and Giles 1947) has been extensively studied, making it the best characterized Gram-positive bacterium. B. amyloli- quefaciens was first isolated in 1943 and the suggested distinction from B. subtilis (Welker and Campbell 1967) is now well documented. B. amyloliquefaciens has been given separate species status and its name has been in- cluded on the approved lists of bacterial names (Priest et al. 1987).

Industrial uses

Bacilli are widely used by the fermentation industry. Well-known examples are B. subtilis, B. amyloliquefa- ciens, B. lichen~ormis, B. alkalophilus, B. lentus and B. thuringiensis. For a recent review on the biotechnology of bacilli see Priest (1990).

Safety aspects

In general, B. subtilis is considered an opportunistic mi- croorganism with no pathogenic potential to humans. However, B. subtilis is virtually ubiquitous and it is

therefore inevitable that it sometimes may be found in association with other microorganisms in infected hu- mans, but only patients treated with immunosuppres- sive drugs appear to be susceptible to infection with this otherwise harmless microorganism (Doyle et al. 1985). We have attempted to collect all pertinent refer- ences reporting such cases and to analyse whether B. subtilis can cause human disease. We mainly refer to cases described after 1970 as confusion between B. cer- eus and B. subtilis existed in diagnostic laboratories be- fore that time (Gordon 1973).

B. subtilis, as well as other Bacillus species, is an im- portant occupant of most environments. A survey by Finch et al. (1978) of the bacterial flora at different sites in 21 homes showed that Bacillus species were present at all of 17 sites in the kitchen and all of 16 sites in the bathroom. Together with Micrococceae, Bacillus spp. were the most frequent organisms isolated. This is probably due to the common occurrence in soil of ba- cilli combined with their ability to produce spores.

Infections

Several authors have noted an increased frequency of registration of infections with Bacillus species (Logan 1988; Kramer and Gilbert 1989). As stated by Logan (1988), this might be associated with improved bacteri- ological techniques and the increasing number of se- verely debilitated patients, for example those who are immunologically compromised.

The literature describing human infections with B. subtilis has been collected from database searches and from our collection of references on Bacillus pathoge- nicity. The search resulted in less than ten relevant arti- cles describing approximately 50 cases of putative B. subtilis infections. Note that this figure is extremely low considering the total number of reports on bacterial in- fections. Almost all cases were related to drug abuse or occurred in severely debilitated patients.

Druo abuse. In drug abusers Tuazon et al. (1979) de- scribed four incidents of endocarditis (i.e. inflamma- tions of the heart). B. cereus was isolated in all cases. Relier (1973) describes one case of endocarditis caused by B. subtilis in a drug abuser.

Infections of drug abusers by bacilli are related to the fact that narcotics are often contaminated by bacilli. Thus, the presence of Bacillus species in narcotics for intravenous administration has been examined. Sham- suddin et al. (1982) investigated 49 heroin samples and found 20 to be contaminated. Of these 13 were contam- inated by Bacillus spp. In a separate study, 47% of the injection utensils and 32% of heroin samples were found to be contaminated by Bacillus species (Weller and Nicholson 1979).

Debilitated patients. Ihde and Armstrong (1973) re- ported on 12 cases of Bacillus spp. infections during a 5-year period from 1966 to 1971. The patients suffered from malignant cancer diseases. Ten of the cases were

described as B. subtilis infections, but as the data were collected before B. cereus and B. subtilis were clearly distinguished from each other, the diagnosis may well be erroneous. Pennington et al. (1976) described two cases of B. subtilis infection in two patients suffering from blood cancer. B. subtilis was isolated from lung and brain tissue.

In a retrospective examination of cases of Bacillus spp. isolated from blood samples at a hospital with a large proportion of immunosuppressed patients, Cot- ton et al. (1987) analysed 17 cases from a 9.5-year peri- od. Fourteen of the patients had chronic venous cathet- ers and B. subtilis was not found in any of the blood samples.

Kiss et al. (1988) reported on 21 B. subtilis bacterem- ias in patients all suffering from debilitating diseases. The treatment of the primary disease in all patients in- cluded insertion of intravenous catheters, lumbar punc- ture or other interventions, which may have introduced the organism to sensitive tissue.

Richard et al. (1988) described 11 cases of Bacillus bacteremias of which B. subtilis was isolated in eight patients. Four of these suffered from cancer diseases and four others had head trauma, stroke or had under- gone surgery. A routine of using B. subtilis culture as a non-specific support for a stable gastrointestinal flora was suspected of being responsible for the infections.

Local infection. Infections of the eye by B. cereus has caused irreversible loss of sight (Shamsuddin et al. 1982). According to literature after 1970, however, B. subtilis seems not to be the agent of infections of the internal eye. Donzis et al. (1988) reported on a case of B. subtilis eye infection related to contamination of contact lenses. Jonas et al. (1981) reported on one case of infection in the shin-bone of a 1-year-old child caused by a splinter in the growth plate of the bone.

Food poisoning. B. cereus is well-established as a cause of food poisoning accounting for 1-23% of the reported foodborne illnesses in humans (Kramer and Gilbert 1989). B. subtilis has been isolated in some cases of food poisoning, but the number of episodes is low. Thus, Kramer and Gilbert (1989) reported on only 49 episodes in the UK in the period 1975-1986. Exact and reliable figures are difficult to obtain, since B. cereus sometimes may have been classified as B. subtilis. As a consequence, there are very few examples of B. subtilis as the confirmed cause of food poisoning.

B. amyloliquefaciens. B. amyloliquefaciens has not ap- peared in any of the cited papers dealing with Bacillus sp. as infectious organisms. A search in databases for references on B. amyloliquefaciens infections or intoxi- cations revealed no such cases, probably because Gor- don et al. (1973) considered B. subtilis and B. amyloli- quefaciens synonymous.

Recombinant strains Conclusion

Since the discovery of plasmids that are able to repli- cate in B. subtilis (Ehrlich 1977), B. subtilis 168 has been used as a host for cloning DNA of both prokaryotic and eukaryotic origin. Considering B. subtilis harmless, the National Institute of Health (US) has exempted sporulation-deficient strains from its Guidelines for Re- search Involving Recombinant Molecules (May 7, 1986). On August 24, 1987, the NIH modified the Guid- elines (Appendix C-IV) such that the physical contain- ment of large-scale fermentation experiments involving sporulation-deficient recombinant B. subtilis does not need to be greater than for the unmodified host.

Permission to produce enzymes from recombinant B. subtilis strains have been given in the US, Japan, and Denmark and the Danish Ministry of Health has issued an environmental certification stating that a recombi- nant B. subtilis product ion strain comply with the OECD recommendations on Good Industrial Large Scale Practice organisms (OECD 1986).

Bielecki et al. (1990), described cloning of the struc- tural gene for Listeria monocytogenes haemolysin, hlyA, into an asporogenic B. subtilis strain. The recombinant, in contrast to the host strain, was able to grow in vitro in the cytoplasm of macrophage-like cells after being internalized. However, the recombinant was absolutely avirulent after intervenous injection in mice and thus did not display any pathogenic properties in vivo. This is in accordance with the general belief that pathoge- nicity is a multifactorial property.

The FDA (Food and Drug Administration of the US) may grant products the status of being "General ly Recognized As Safe" (GRAS). Evidence, which in FDA's opinion may lead to this conclusion is published for public comment as a GRAS petition, which even- tually may lead to clearance as GRAS.

In a GRAS petition by CPC International (1986), the company reviewed the pathogenicity and toxicogenicity of B. subtilis. A search covering the period 1907-1983 failed to disclose a single report demonstrating that B. subtilis can be the etiological agent of diseases in man or animals. In the GRAS petition it is noted that al- though B. subtilis strains have sometimes been reported to be implicated in food poisoning, the reports are spe- culative and in no cases were confirmatory toxicologi- cal studies conducted. In the same GRAS petition, spe- cific toxicological studies showed that an a-amylase from B. stearothermophilus produced by a recombinant B. subtilis is safe for use in food. In another GRAS pe- tition Enzyme Bio-Systems (1988) demonstrated the sa- fety of a B. rnegaterium amylase produced by a recom- binant B. subtilis. Finally, a GRAS petition from Novo Laboratories (1990) included safety data on a maltog- enic amylase produced by a recombinant B. subtilis. Andersen et al. 1987 published a safety study on the toxicological and mutagenic potential of the same en- zyme.

No case demonstrating invasive properties of Bacillus subtilis or B. amyloliquefaciens has been described but in a few cases, B. subtilis has been found associated with drug abusers or severely debilitated patients. Thus there is no evidence of any pathogenic potential of B. subtilis to humans in general. B. subtilis has been asso- ciated with some cases of food poisoning which in part may be due to misclassification of B. cereus. Thus there are very few examples of B. subtilis strains as confirmed causes of food poisoning. We conclude that B. subtilis is a safe host for the production of harmless products.

Acknowledgement. We are grateful to Dr. Fergus Priest, Edin- burgh, for critical comments on the manuscript.

References

AMFEP (Association of Microbial Food Enzyme Producers) (1990) Regulatory aspects of food enzymes produced by re- combinant microorganisms. Association of Microbial Food Enzyme Producers, Bruxelles

Andersen JR, Diderichsen BK, Hjortkjaer RK, Boer AS de, Boot- man J, West H, Ashby R (1987) Determining the safety of mal- togenic amylase produced by rDNA technology. J Food Pro- tect 50:521-526

Bielecki J, Youngmans P, Connelly P, Portnoy DA (1990) Bacillus subtilis expressing a haemolysin gene from Listeria monocyto- genes can grow in mammalian cells. Nature 345:175-176

Burkholder PR, Giles NH (1947) Induced biochemical mutants in Bacillus subtilis. Am J Bot 34:345-348

Cotton D J, Gill VJ, Marshall D J, Gress J, Thaler M, Pizzo PA (1987) Clinical features and therapeutic interventions in 17 cases of Bacillus bacteriemia in an immunosuppressed patient population. J Clin Microbiol 25:672-674

CPC International (1986) GRAS Petition 4G0293 proposing that alpha-amylase from a strain of Bacillus subtilis (ATCC 39 705) containing the gene for alpha-amylase from B. stearothermo- philus inserted by recombinant DNA techniques be affirmed as GRAS as a direct human food ingredient. Notice of Filing. Federal Register 51:10571

Diderichsen Bet al. (1990) The conditions for contained use of genetically engineered microorganisms and cells for industrial production in Europe. Biotechnol Forum Eur 7:484-485

Djien KS, Hesseltine CW (1979) Tempe and related foods. In: Rose AH (ed) Economic microbiology, vol 4. Academic Press, London, pp 116-140

Donzis PB, Mondino B J, Wiessman BA (1988) Bacillus keratitis associated with contaminated contact lens care systems. Am J Ophthalmol 105:195-197

Doyle RJ, Keller KF, Ezzell JW (1985) In: Lennette EH, Ballows A, Hausler WJ, Shadomy HJ (eds) Manual of clinical micro- biology, 4th edn. American Society for Microbiology, Wash- ington, D. C., pp 211-215

Ehrlich SD (1977) Replication and expression of plasmids from Staphylococcus aureus in Bacillus subtilis. Proc Natl Acad Sci USA 74:1680-82

Enzyme Bio-Systems (1988) GRAS Petition 7G0328 proposing that alpha-amylase from a strain of Bacillus subtilis containing a Bacillus megaterium alpha-amylase gene be affirmed as GRAS as a direct human food ingredient. Notice of Filing. Federal Register 53 : 16191

Finch JE, Prince J, Hawksworth M (1978) A bacteriological sur- vey of the domestic environment. J Appl Bacteriol 45:375- 364

Gordon RE, Haynes WC, Pang CH-N (1973) The genus Bacillus.

In: Gordon RE, Haynes WC, Pang CH-N (eds) Agricultural Handbook ~ no. 427. US Department of Agriculture, Washing- ton, D. C., pp 23-25

IFBC (International Food Biotechnology Council) (1990) Bio- technologies and food: assuring the safety of foods produced by genetic modification. Regul Toxicol Pharmacol 12: Special Issue

Ihde DC, Armstrong D (1973) Clinical spectrum of infection due to Bacillus species. Am J Med 55:839-845

Jonas M, Cunha BA, Greensher J (1981) Bacillus subtilis osteomy- celitis. South Med J 74:1421-1422

Kiss T, Gratwohl A, Frei R, Osterwalder B, Tichelli A, Speck B (1988) Bacillus subtilis Infectionen. Schweiz Rundsch Med 77:1219-1223

Kramer JM, Gilbert RJ (1989) Bacillus cereus and other Bacillus species. In: Doyle MP (ed) Foodborne bacterial pathogens. Dekker, New York, pp 21-70

Logan NA (1988) Bacillus species of medical and veterinary im- portance. J Med Microbiol 25:157-165

National Academy of Sciences (1987) Introduction of recombi- nant DNA-engineered organisms into the environment: key is- sues. National Academy Press, Washington, DC

Novo Laboratories (1990) GRAS Petition 7G0326 proposing that maltogenic amylase enzyme preparation derived from a gene- tically modified Bacillus subtilis be affirmed as GRAS as a di- rect human food ingredient. Federal Register 55:9772-9773

OECD (1986) Recombinant DNA safety considerations. Paris Pariza MW, Foster EM (1983) Determining the safety of enzymes

used in food processing. J Food Prot 46:453-468 Pennington JE, Gibbons ND, Strobeck JE, Simpson GL, Myero-

witz RL (1976) Bacillus species infections in patients with hae- matologic infections. J Am Med Assoc 235:1473-1474

Priest FG (1990) Products and applications. In: Harwood CR (ed) Bacillus. Plenum, New York, pp 293-320

Priest FG, Goodfellow M, Shute LA, Berkeley RCW (1987) Bacil- lus amyloliquefaciens sp. nov., nom. rev. Int J Syst Bacteriol 37:69-71

Relier LB (1973) Endocarditis caused by Bacillus subtilis. Am J Clin Pathol 60:714-718

Richard V, Auwera P van der, Snoeck R, Daneau D, Meunier F (1988) Nosocomial bacteremia caused by Bacillus species. Clin Microbiol Infect Dis 7:783-785

Shamsuddin D, Tuason CU, Levy C, Curtin J (1982) Bacillus cer- eus phanophthalmitis: source of the organism. Rev Infect Dis 4:97-103

Tuazon CV, Murray HW, Levy C, Solny MN, Curtin JA, Sheagren JN (1979) Serious infections from Bacillus sp. J Am Med Assoc 241 : 1137-1140

Welker NE, Campbell LL (1967) Unrelatedness of Bacillus amylo- liquefaciens and Bacillus subtilis. J Bacteriol 94:1124-30

Weller PF, Nicholson A (1979) The spectrum of Bacillus bacter- emias in heroin addicts. Arch Internal Med 139:293-294